2021-01-05T15:39:22Z
Acceptance in OA@INAF
The infrared-luminous progenitors of high-z quasars
Title
Ginolfi, M.; Schneider, R.; VALIANTE, ROSA; Pezzulli, E.; Graziani, L.; et al.
Authors
10.1093/mnras/sty3205
DOI
http://hdl.handle.net/20.500.12386/29497
Handle
MONTHLY NOTICES OF THE ROYAL ASTRONOMICAL SOCIETY
Journal
483
Number
MNRAS 000,1–10(2018) Preprint 28 November 2018 Compiled using MNRAS LATEX style file v3.0
The infrared-luminous progenitors of high-z quasars
M. Ginolfi
1,2?, R. Schneider
2,3, R. Valiante
1, E. Pezzulli
2,3, L. Graziani
2,
S. Fujimoto
4, and R. Maiolino
5,61INAF/Osservatorio Astronomico di Roma, Via di Frascati 33, 00040 Monte Porzio Catone, Italy 2Dipartimento di Fisica, Sapienza Universit`a di Roma, Piazzale Aldo Moro 5, I-00185, Roma, Italy 3INFN, Sezione di Roma I, P.le Aldo Moro 2, 00185 Roma, Italy
4Institute for Cosmic Ray Research, The University of Tokyo, Kashiwa, Chiba 277-8582, Japan 5Cavendish Laboratory, University of Cambridge, 19 J.J Thomson Ave., Cambridge CB3 0HE, UK 6Kavli Institute for Cosmology, University of Cambridge, Madingley Road, Cambridge CB3 0HA, UK
Accepted XXX. Received YYY; in original form ZZZ
ABSTRACT
Here we explore the infrared (IR) properties of the progenitors of high-z quasar host galaxies. Adopting the cosmological, data constrained semi-analytic model GA-METE/QSOdust, we simulate several independent merger histories of a luminous quasar at z ∼ 6, following black hole growth and baryonic evolution in all its progenitor galaxies. We find that a fraction of progenitor galaxies (about 0.4 objects per single luminous quasar) at 6.5 < z < 8 has an IR luminosity of LIR > 1013 L
(hyper-luminous IR galaxies; HyLIRGs). HyLIRGs progenitors reside in the most massive halos, with dark matter (DM) masses of MDM∼ 1012.5− 1013M . These systems can
be easily observed in their ∼ 1 mm-continuum emission in a few seconds of integration time with the Atacama Large Millimeter/submillimeter Array (ALMA), and at least 40% of them host nuclear BH activity that is potentially observable in the soft and hard X-ray band. Our findings are in line with recent observations of exceptional mas-sive DM halos hosting HyLIRGs at z ∼ 7, suggesting that z ∼ 6 luminous quasars are indeed the signposts of these observed rare peaks in the high-z cosmic density field, and that massive IR-luminous galaxies at higher z are their natural ancestors. Key words: Galaxies: evolution, high-redshift; quasars: general, supermassive black holes; infrared: galaxies; submillimetre: galaxies
1 INTRODUCTION
Observations of dusty star-forming galaxies (DSFGs) at high-redshift, often selected at (sub-)millimeter wavelengths (submillimeter galaxies, or SMGs;Smail et al. 1997;Hughes et al. 1998; Chapman et al. 2005; Hayward et al. 2013), can provide unique insights into our understanding of the early formation of massive galaxies (see e.g., Blain et al. 2002;Casey et al. 2014for reviews). Because of their large dust content, DSFGs emit most of their luminosity at infrared (IR) wavelengths (8 − 1000 µm) and are typically considered ultra-luminous IR galaxies (ULIRGs), reaching IR luminosities (LIR) of LIR > 1012 L (Kov´acs et al.
2006; Coppin et al. 2008; Hayward et al. 2011). Spectro-scopic surveys targeting the redshift distributions of SMGs indicate that the bulk of the DSFGs population peaks at z ∼ 2 − 4 (e.g.,Chapman et al. 2005;Strandet et al. 2016), encompassing the peaks of supermassive black hole (SMBH)
? E-mail: michele.ginolfi@inaf.it
accretion (Cattaneo & Bernardi 2003;Hopkins et al. 2007) and cosmic star formation activity (Madau & Dickinson 2014). However, a significant tail of higher redshift DSFGs appears to be already in place at z > 5 (e.g.,Riechers et al. 2013, 2017;Walter et al. 2012; Weiß et al. 2013;Strandet et al. 2017). The latter are often found to be extreme hyper-luminous IR galaxies (HyLIRGs), reaching infrared luminosities of LIR > 1013 L and star formation rates (SFRs) exceeding 1000 M yr−1, likely tracing the result of strong dynamical interactions (i.e., major mergers) and intense gas accretion events occurring in the densest regions of the early Universe (e.g.,Riechers et al. 2011,2017;Ivison et al. 2011;Capak et al. 2011;Oteo et al. 2016;Pavesi et al. 2018).
Recently,Marrone et al.(2018) reported high-resolution At-acama Large Millimeter/submillimeter Array (ALMA) ob-servations of SPT0311-58, a IR-luminous system at z = 6.9, originally identified in the 2500 deg2 South Pole Telescope (SPT) survey (Carlstrom et al. 2011). ALMA reveals this source to be a pair of interacting extremely massive and
IR-c
2018 The Authors
luminous star-bursting galaxies: SPT0311-58W, a HyLIRG with LIR= 3.3 ± 0.7 × 1013L and a SFR ∼ 2900 M yr−1, and SPT0311-58E, a ULIRG with LIR= 4.6 ± 1.2 × 1012L and SFR ∼ 540 M yr−1. Using different proxies,Marrone
et al. (2018) estimated the DM halo mass hosting this system to be MDM = (1.4 − 7) × 1012 M , showing that SPT0311-58 marks a rare peak in the cosmic density field at this early cosmic time and it lies close to the exclusion curve predicted by the current structure formation paradigm. Many models of SMBH-galaxy co-evolution suggest an evolutionary link between ULIRGs and quasars (quasars;
Sanders et al. 1988a,b; Silk & Rees 1998; Springel et al. 2005;Di Matteo et al. 2005;Hopkins et al. 2006,2008, and
Alexander & Hickox 2012for a review). ULIRGs may repre-sent the initial, heavily obscured, stages of quasars, which, after shedding the surrounding dust through energetic galaxy-scale outflows (e.g., Maiolino et al. 2012; Harrison et al. 2012; Fan et al. 2018), evolve into a ultraviolet (UV)/optical bright unobscured phase.
In this work, we explore the hierarchical merger histories of z ∼ 6 quasars to investigate their connection with ULIRGs and HyLIRGs. To this aim, we use GAMETE/QSOdust (here-after GQd,Valiante et al. 2014,2016), a semi-analytic model tested to reproduce the observed properties of a sample of high-z quasars (Valiante et al. 2014). GQd consistently follows the formation and evolution of nuclear black holes (BHs) and their host galaxies, accounting for star formation, interstellar medium (ISM) chemical evolution (metals and dust) and mechanical feedback, along different merger histories of the parent DM halo. Therefore, GQd is a suitable tool to (i) investigate the physical properties of the ancestors at z ∼ 7 − 9 of luminous quasars at z ∼ 6; (ii) statistically quantify the number density of expected ULIRG and HyLIRG progenitors, and (iii) probe whether z ∼ 6 luminous quasars can be the signposts of the rare peaks in the cosmic density field observed at z ∼ 7 (i.e.,
Marrone et al. 2018).
The paper is organized as follows. In Section 2, we briefly describe the GQd model. Section3and4show the results of this work and a discussion of their implications. Throughout the paper, we assume a ΛCDM cosmology with Ωm= 0.24, ΩΛ= 0.76, Ωb= 0.04, and H0= 73 km s−1 Mpc−1.
2 MODEL DESCRIPTION
GQd1 is a semi-analytic, data-constrained model aimed at studying the formation and evolution of high-redshift quasars and their host galaxies in a cosmological frame-work. In this section we summarize the main features of GA-METE/QSOdust, referring the reader toValiante et al.(2011,
2014,2016) for a thorough description of the model. GQd has been tested to reproduce the observed properties of quasars at z > 5, such as the BH mass, the molecular gas and dust masses in the ISM (Valiante et al. 2014). Following
1 The model GAMETE (GAlaxy MErger Tree and Evolution) was
originally developed to study the formation and cosmological evo-lution of local, Milky-Way like galaxies (Salvadori et al. 2007, 2008).
Table 1. Observed and inferred properties of the quasar SDSS J1148+5251 at z = 6.42.
(a)L
FIR (b)Lbol (a)SFR (a)Mdust (c)MBH
[1013L ] [1014L ] [103M yr−1] [108M ] [109M ]
2.2 ± 0.33 1.36 ± 0.74 2.0 ± 0.5 3.4+1.38−1.54 4.9 ± 2.5
(a) The values of LFIR, SFR, and Mdusthave been computed by Valiante et al.(2011,2014).
(b) The bolometric luminosity, Lbol, is estimated from the
ob-served flux at 1450 ˚A (Fan et al. 2003) using the bolometric correction byRichards et al.(2006).
(c) The black hole mass, MBH, is estimated from the MgII
dou-blet and the λ = 3000 ˚A continuum (De Rosa et al. 2011).
Valiante et al.(2016), in this work we focus on the evolu-tion of a SDSS J1148+5251 (hereafter J1148)-like quasar at z = 6.4, one the best studied high-z luminous quasars2(e.g.,
Willott et al. 2003;Walter et al. 2004;Maiolino et al. 2005;
Cicone et al. 2015), and we use it as a prototype for the general class of luminous z ∼ 6 quasars.
The hierarchical merger history
We first reconstruct different hierarchical merger histories (merger trees) of a MDM = 1013M DM halo hosting a J1148-like quasar at z = 6.4, using a binary Monte Carlo algorithm accounting for mass infall (e.g., Volonteri et al. 2003), based on the Extended Press-Schechter theory (e.g.,
Lacey & Cole 1993). This approach enables us to produce a large (statistically meaningful) number of random, semi-analytical merger trees of the same DM halo, simultane-ously resolving the first collapsed objects. The DM reso-lution mass, Mres, is chosen to resolve the first collapsed objects where the gas is able to cool and form stars, i.e., the so-called mini-haloes, with masses of Mh∼ 106M and virial temperatures of Tvir < 104 K (see Bromm 2013 for a review). Given this requirement, we simulate the merger trees adopting a Mres of:
Mres(zi) = 10−3Mh(z0)
1 + zi 1 + z0
−7.5
, (1)
where Mh(z0) is the host DM halo at redshift z0= 6.4. Mini-haloes dominate the halo mass spectrum at very high-z; at z . 14 the merger tree halo population is dominated instead by Lyα-cooling halos, namely DM halos with Tvir ≥ 104K (seeValiante et al. 2016for details).
By construction, the entire population of halos along the merger trees, that we call progenitors, contribute to assem-bling the simulated quasar host galaxies. However, when specified within the text, we will restrict the analysis to mas-sive progenitors, hosted by DM halos with MDM> 1011M . This mass-selection, as discussed in Sec.3, reflects our in-terest of studying the most IR-luminous galaxies in the sim-ulation.
2 Absolute AB magnitude of the continuum in the rest-frame at
The IR-luminous progenitors of high-z quasars
3
BH growth and baryonic evolution
Merger trees are used as an input to reconstruct the formation and evolution of the J1148-like quasar and its host galaxy through cosmic time. The baryonic evolution inside each progenitor galaxy is regulated by processes of star formation, BH growth and feedback, and followed consistently through the cosmic mass assembly.
BHs growth is driven by:
- mergers: we assume that in major mergers3 the BHs
present in the nuclei of the two interacting galaxies coalesce, forming a new more massive BH. In minor mergers, the merger time-scale of the two BHs is of the order of the Hubble time or longer (see e.g., Tanaka & Haiman 2009). The less massive BH of the merging pair is assumed to remain as a satellite and we do not follow its evolution. - gas accretion: the accretion rate is described by a modi-fied4 Bondi-Hoyle-Lyttleton (BHL) formula, capped at the Eddington rate (see Valiante et al. 2014 for details). We find that, while BH growth is mostly driven by mergers at very high-z (i.e., z & 11), gas accretion is the dominant growth mode at lower redshifts (Valiante et al. 2016). At each snapshot of the merger tree, stars in the pro-genitor galaxies form according to a SFR proportional to the available gas mass, with an efficiency that is enhanced during major mergers (see Valiante et al. 2014). Following star formation, supernovae (SNe) and asymptotic giant branch stars (AGB) progressively enrich the ISM of each galaxy with metals and dust according to their specific mass and metallicity-dependent yields and to their evolutionary time-scales. We follow the life-cycle of metals and dust including physical prescriptions for dust processing in a two-phase ISM: SN shocks can destroy dust grains in the hot, diffuse medium while dust grains can grow in mass by accreting gas-phase heavy elements in the cold medium (see Valiante et al. 2014and de Bennassuti et al. 2014, for details). A fraction of the energy released by SN explosions (0.2%) and BH accretion (0.25%) is converted into kinetic energy of the gas in the host galaxy, driving gas outflows in the form of winds. While these efficiencies are lower than what generally adopted in hydrodynamical simulations5 the predicted strength of mechanical feedback is in good agreement with observations of the outflowing gas in J1148 (Maiolino et al. 2005, 2012; Valiante et al. 2012;Cicone et al. 2015; see alsoBischetti et al. 2018, who carry out a stacking analysis of a sample of 48 quasars at 4.5 < z < 7.1 detected by ALMA in the [CII] 158 µm line, and find an outflow kinetic power that is ∼ 0.1% of the AGN luminosity).
3 A major merger occurs when two DM halos with mass ratio
(less massive over the most massive) µ > 1/4 coalesce.
4 We re-scaled by a factor α
BH= 50, to account for the higher
central densities around accreting BHs (e.g., Di Matteo et al. 2005).
5 In one of the most popular descriptions of AGN-driven winds in
hydrodynamical simulations, 5% of the radiation energy is ther-mally coupled to the surrounding gas (Di Matteo et al. 2005). This difference may be due to the complex interaction of the outflow with the inhomogeneous ISM and the effects of radiative losses that can not be captured by our simple semi-analytical model.
Table 2. Compilation of the k0, λ0 and β parameters, defining
the opacity coefficient per unit dust mass in the different models of dust composition used in this work.
Model k0 λ0 β
(reference) [cm2/gr] [µm] [β] Bertoldi et al.(2003) 7.5 230 1.5 Robson et al.(2004) 30 125 2.0 Beelen et al.(2006) 0.4 1200 1.6 Weingartner & Draine(2001)(a) 34.7 100 2.2 Bianchi & Schneider(2007) 40 100 1.4 THEMIS -Jones et al.(2017) 6.4 250 1.79
AC -Galliano et al.(2011) 16 160 1.7 (a) Fit to the model for the Small Magellanic Cloud in the spec-tral range [40 − 200] µm.
3 THE IR PROPERTIES OF MASSIVE
PROGENITORS OF HIGH-Z QUASAR HOST GALAXIES
We calculate the IR luminosity of each galaxy by assuming that in the Rayleigh-Jeans part of the spectrum (in the limit of small frequencies, that is hν kBT ), dust radiates as a ‘grey-body’ with an opacity coefficient per unit dust mass, kd(ν) = k0 (ν/ν0)β, where ν0, k0, β depend on the adopted model of dust. Integrating over the IR wavelength range 8 − 1000 µm (300 − 38000 GHz), we compute the corresponding LIRas,
LIR= 4πMdust Z
kd(ν) B(ν, Tdust) dν, (2)
where B(ν, Tdust) is the Planck function for a dust tempera-ture Tdust. We estimate LIR, averaging over 7 different mod-els of dust composition6 (each of them providing a unique combination of the parameters k0, ν0, β; see Table2) and we vary the dust temperature7in the range Tdust= [35 − 55] K.
Fig. 1 and Fig. 2 show respectively (i) the redshift distribution of simulated DM halo masses and (ii) the IR luminosities of massive progenitors (MDM > 1011M ) of J1148-like quasar host galaxies at z ∼ 6.4, in all the simulated merger trees. Fig. 1 shows that, although the bulk of the mass distribution of massive progenitors is concentrated in the mass range 11 < log(MDM/M ) < 12 (especially at z ∼ 8; see percentile levels in the density plot), IR-luminous galaxies with LIR> 1012L occupy DM halos with MDM> 1012M , at least in the redshift interval 6.4 < z < 7.5. In particular, we find that galaxies with HyLIRGs-like luminosity (LIR > 1013L , consistent with the IR luminosity measured in SPT0311-58W; Marrone et al. 2018) reside in the most massive halos, with masses of
6 FollowingValiante et al.(2011) we use models of dust
compo-sition fromBertoldi et al.(2003), Robson et al. (2004), Beelen et al.(2006),Weingartner & Draine(2001),Bianchi & Schneider (2007). In addition, we also consider the more recent ‘AC’-model byGalliano et al.(2011) and ‘THEMIS’ byJones et al.(2017) (see the review byGalliano et al. 2017for a discussion).
7 Warm interstellar dust, associated to starburst regions,
dom-inates the emission in the rest-frame far-IR (Dunne et al. 2000; Wang et al. 2008;Valiante et al. 2016).
Figure 1. Density plot showing the MDM distribution of DM
halos hosting massive progenitor galaxies (MDM> 1011M ) of a
J1148-like quasar host galaxy (red point), in the redshift interval 6.4 < z < 9. The density distribution is based on the total statis-tics of 10 different merger trees (see Sec.2); the percentile levels are marked by the filled colour contours and labeled in the figure. The red dashed line and points represent the redshift evolution of the averaged MDMof all massive progenitors (MDM> 1011M ).
The gold and green lines and points represent the redshift evo-lution of the average MDMof halos hosting massive progenitors
with LIR> 1012L (ULIRGs) and LIR > 1013L (HyLIRGs),
respectively. Error-bars are indicative of the 1σ standard devi-ation calculated in equispaced redshift bins (the slight offset of the bins position for red, gold and green points has been intro-duced to avoid overlapping and help visualization). The orange point represents the DM halo mass hosting the HyLIRG system SPT0311-58, at z = 6.9 (Marrone et al. 2018). The blue solid and magenta dashed lines represent the 1σ exclusion curves, i.e., the most massive halos that are expected to be observable within the whole sky and a 2500 deg2 area (the size of the SPT survey in
which SPT0311-58 was identified) respectively, as a function of redshift (Harrison & Hotchkiss 2013;Marrone et al. 2018).
about MDM∼ 1012.5− 1013M , consistent with the dynam-ical mass inferred for the interacting system SPT0311-58 (see orange hexagon in Fig.1).
These findings indicate that exceptionally massive DM halos hosting HyLIRGs, as the ∼ 1012.6M
halo hosting SPT0311-58 at z ∼ 7, can be found within the family tree of UV/optical bright quasars at z ∼ 6.
In the next paragraphs we explore the expected physical properties of the IR-luminous progenitors of high-z lumi-nous quasar host galaxies, as well as their frequency (i.e., the expected number per quasar) and their observability in the mm and X-ray bands.
3.1 How many IR-luminous progenitor galaxies do we expect for each z ∼ 6 luminous quasar? In the top panel of Fig. 2 we show the predicted num-ber of IR-luminous progenitor galaxies in the redshift in-terval 6.4 < z < 9 expected for each luminous J1148-like
10
9
8
7
6
Figure 2. Bottom panel: redshift distribution of the LIRin
mas-sive progenitors (MDM > 1011M ) of a J1148-like quasar host
galaxy, in the redshift interval 6.4 < z < 9, using the total statis-tics of 10 independent merger trees. The distribution is colour coded depending on the galaxy ISM dust mass. The black dashed lines represent the density contours with the corresponding per-centiles labeled in the figure. The grey (magenta) filled regions define the LIR-range of ULIRGs (HyLIRGs). The red point
rep-resents the LIR of SPT0311-58W, the most massive galaxy in
the interacting system SPT0311-58 (Marrone et al. 2018). Top panel: average number of ULIRGs (grey points) and HyLIRGs (magenta points) at 6.4 < z < 9, expected for each luminous quasar at z ∼ 6. Error-bars are indicative of the 1 σ standard deviation calculated over 10 different merger trees.
quasar at z ∼ 6.4. Each high-z luminous quasar has, on av-erage, about 2.8 ULIRG progenitors with LIR > 1012L , and about 0.4 HyLIRG progenitors with LIR > 1013L , between z ∼ 6 − 8. Recently, Jiang et al. (2016) using a complete sample of 52 quasars at z > 5.7 selected from the Sloan Digital Sky Survey (SDSS), spanning a wide luminos-ity range of −29.0 6 M14506 −24.5, found that the number density of bright quasars with M1450 < −26.7 at z ∼ 6 is (0.4 ± 0.1) Gpc−3 8. Combining this information with our estimate of the number of expected IR-luminous progenitor galaxies per quasar, we predict the spatial number density of HyLIRG (ULIRG) progenitors to be about 0.16 ± 0.08 (1.2 ± 0.6) Gpc−3 in the redshift range z ∼ 6 − 8. Since these values refer to the population of ULIRGs/HyLIRGs found within the merger trees of UV/optical bright quasars at z ∼ 6, they provide lower limits on the effective num-ber density of the overall population of IR-luminous galaxies at z > 6. Potential deviations might be also driven by the effect of the duty cycle on the number counting of high-z
8 This value is consistent with the estimates provided by previous
The IR-luminous progenitors of high-z quasars
5
quasars, resulting in a systematic under-estimation of their number densities in large single-epoch imaging surveys (see e.g., Romano-Diaz et al. 2011). An alternative way to es-timate the number density of the overall population of IR-luminous galaxies at z > 6 comes from our finding that, at 6.5 < z < 8, the most massive DM halos (MDM> 1012.5M ) in the merger trees of simulated quasars are most likely to host HyLIRGs (see Fig.1). Assuming that this finding can be extended to all DM halos in the same mass regime (i.e., also halos that are not expected to evolve in a quasar at z ∼ 6), then the number density of the HyLIRGs population should be consistent with the number density of DM halos with MDM> 1012.5M predicted by the ΛCDM cosmology9 at 6.5 < z < 8, i.e., about 2 Gpc−3.
Fig. 2 also shows that HyLIRGs are characterized by dust masses & 108.7M already at z ∼ 8−8.5, while ULIRGs have dust masses in the range 107.5< Mdust/M < 109. Despite their large stellar masses, & 70% of the dust mass has formed in the ISM of these galaxies through processes involving some form of grain growth10 (see Valiante et al. 2011, where the maximum dust masses obtained from stellar sources, e.g., SNe and AGB stars, is calculated as a function of Mstar, under the conservative assumption that dust grains are not destroyed by interstellar shocks).
3.2 The physical properties of high-z HyLIRGs In Fig. 3 we show the redshift, stellar mass (Mstar), gas mass (Mgas), SFR, dust mass (Mdust), metallicity (Z) and BH mass (MBH) distributions of HyLIRG progenitors (ma-genta histograms) of high-z quasars. We find that HyLIRG progenitors are generally highly massive (i.e., Mstar∼ 1011− 1012M
; Fig.3.b), gas rich (i.e., Mgas∼ 1010.5− 1011.5M ; Fig.3.c) systems. They experience strong starbursts, with SFRs ∼ 102− 104M
yr−1(see Fig.3.d), enabling them to reach super-solar metallicities (see Fig.3.e) and large dust masses (i.e., Mdust∼ 109M ; Fig.3.f) already at z ∼ 6.5 − 7.5. The same properties are also shown for galaxies experi-encing mergers with HyLIRG companions along the merger trees (light green histograms). Those galaxies are found to have different properties than HyLIRGs (note the differ-ent x-axis range between magdiffer-enta and green histograms in Fig. 3). In particular, we find that the bulk of the popu-lation of interacting companions is generally less massive (the Mstar distribution peaks at Mstar∼ 109M ), less star-forming (the SFR distribution peaks at SFR ∼ 5 M yr−1) and less chemically evolved (i.e., sub-solar metallicities and Mdust ∼ 106.5M ). However, we find that the tails of the distributions of interacting companions extend toward the HyLIRG regime, as shown by the intersections between the light green histograms and the magenta lines in Fig.3. The existence of this channel of interaction is crucial from the point of view of evolutionary models of SMBH-galaxy co-evolution: in fact, gas-rich major mergers are thought to be
9 The ΛCDM-predicted number density of DM halos was
ob-tained using HMFcalc , publicly available athttp://hmf.icrar.org/ (seeMurray et al. 2013).
10 We note that the efficiency and even the physical nature of
grain growth is still an active subject of study (e.g.,Zhukovska et al. 2016;Ceccarelli et al. 2018;Ginolfi et al. 2018).
an efficient mechanism to deliver cold gas at the centre of the resulting halo and therefore fuel the growth of the SMBH and trigger AGN activity (Di Matteo et al. 2005;Hopkins et al. 2008;Pezzulli et al. 2016).
3.3 The observability of high-z HyLIRGs with ALMA
By assuming an optically thin, τd(λ) << 1, rest-frame far-IR emission, the intrinsic flux observed in a given band Fν can be estimated as,
Fν=
Mdustkd(ν) B(ν, Tdust) (1 + z) d2
L(z)
, (3)
where d2L(z) is the luminosity distance of a source at given redshift.
However, it is well known that at high redshift the cosmic microwave background (CMB) is hotter and brighter. Its temperature, TCMB(z) = TCMBz=0 (1 + z), at z > 5 approaches the temperature of the cold dust, decreasing the contrast of the intrinsic rest-frame IR emission against the CMB. To account for this effect, we adopted the results byda Cunha et al.(2013), who provided general correction factors to es-timate what fraction of the dust emission can be detected against the CMB as a function of frequency, redshift and temperature, Fνagainst CMB Fintrinsic ν = 1 −B(ν, TCMB(z)) B(ν, Tdust(z)) . (4)
It follows from Eq.4that the flux-CMB contrast increases at higher Tdust (see also the discussion in da Cunha et al.
2013). This is the case for our simulated HyLIRGs, where the interstellar dust in the warm phases of the ISM is heated by the vigorous starburst activity, reaching the high temperature values adopted in this paper (Tdust & 40 K, see also Dunne et al. 2000; Wang et al. 2008; Valiante et al. 2016). Indeed we find that a large fraction, namely 70 − 80%, of the intrinsic rest-frame IR emission can be detected against the CMB.
Fig. 4 shows the distribution (both averaged values and dispersions) of the rest-frame IR fluxes (CMB-corrected ) for HyLIRG progenitors at 6.4 < z < 9. Fluxes are calculated in the ALMA Bands [8, 7, 6, 4, 3], respectively centered at observed-frame wavelengths [0.7, 0.95, 1.25, 2, 3] mm, and range from a minimum value of 0.2 mJy at 3 mm (Band 3), to a maximum value of 50 mJy at 0.7 mm (Band 8). We calculated the on-source integration times11 needed to
detect the estimated continuum fluxes with a signal-to-noise of S/N = 10. We find that HyLIRGs progenitors of high-z quasar host galaxies can be easily detected with ALMA with a modest amount of observing time, ranging from few second in Bands 6-7-8, to ∼ 20 minutes in Band 3 (see Fig.4). We note that, although for such short integrations most of the time is dedicated to the instrumental setup and calibrations, the total integration time (including overheads) needed to reach the required sensitivity is less than 30 min in all the Bands.
11 The on-source integration times were obtained using the
ALMA Cycle 6 sensitivity calculator. MNRAS 000,1–10(2018)
Normal ized d istribut ion Normal ized d istribut ion a) b) c) d) e) f) g)
Figure 3. The magenta histograms show the distributions of physical properties (Mstar, Mgas, SFR, Z, Mdust, MBH) and redshift of
HyLIRGs progenitors of z ∼ 6 quasars (namely the galaxies included in the magenta filled region of Fig.2). The light green histograms show the same distributions for systems experiencing mergers with HyLIRGs progenitors. The overlaid magenta lines indicate the range of values of HyLIRGs properties, revealing clear differences between HyLIRGs and their interacting companions. The overlaid red and dark green solid areas indicate the observed values of the two galaxies composing the interacting system observed byMarrone et al.(2018), i.e., SPT0311-58W and SPT0311-58E respectively. The width of the area is representative of the uncertainty on the measurements.
20 sec (10 σ) 8 sec (10 σ) 10 sec (10 σ) 2 min (10 σ) 21 min (10 σ)
Figure 4. Rest-frame far-IR fluxes of HyLIRG progenitors of a J1140-like quasar host galaxy. Fluxes are calculated in the ALMA Bands [8, 7, 6, 4, 3] (coloured filled regions) respectively centered at observed-frame [0.7, 0.95, 1.25, 2, 3] mm. Magenta points rep-resent the averaged values of the HyLIRGs population, and bars represent their dispersions (the colour intensity traces the flux dis-tribution at any given wavelength bin). The red points represent the fluxes of SPT0311-58W (Marrone et al. 2018) in the ALMA Bands [8, 7, 6, 4, 3]. For fluxes corresponding to the grey bars, located at the low-end of the HyLIRG flux distributions at any wavelength bin, we report the on-source observing time needed by ALMA to reach a S/N = 10 σ.
3.4 The X-ray observability of BH activity in high-z HyLIRGs
As shown in Fig.3.g, we find that the HyLIRG progenitors of luminous high-z quasars, host nuclear BHs with an al-most uniform distribution of masses in the range 105M < MBH < 108M , and a high-mass tail at MBH . 109M (populated by systems in the last steps of the merger trees). Such a broad range of nuclear BH masses results from the diversity of merging and accretion histories predicted by the model. To assess the possibility of detecting BH activity in the nuclei of high-z HyLIRGs, we compared the predicted X-ray luminosity in the (observed-frame) soft ([0.5 - 2] keV) and hard ([2 - 10] keV) X-ray bands with the flux limits of recent surveys.
Following Pezzulli et al.(2017), we modeled the X-ray lu-minosity of accreting BHs considering the primary emission from the hot corona and the reflection component due to the surrounding neutral medium. The first one is parametrized as a power law, Lν∝ ν−Γ+1ehν/Ec, with an exponential cut-off energy of Ec = 300 keV (Sazonov et al. 2004;Yue et al.
2013) and photon spectral index Γ = 0.32 log λEdd+ 2.27, that we assume to depend on the Eddington ratio λEdd = Lbol/LEdd (Brightman et al. 2013). The metallicity depen-dent reflection component is computed using the PEXRAV
The IR-luminous progenitors of high-z quasars
7
[0.5 - 2] keV observed-frame [2 - 10] keV observed-frame [0.2 - 10] keV rest-frame 40 % 38 %a)
b)
c)
Figure 5. a ) Distribution of the intrinsic (magenta histogram) and obscured (red histogram) X-ray luminosities (integrated over the rest-frame soft+hard band, [0.2 -10] keV) produced by the accreting BHs in the HyLIRG progenitors of high-z quasars. b) and c) Distributions of the intrinsic and obscured X-ray fluxes produced by the accreting BHs, in the observed-frame soft (panel b) and hard (panel c) X-ray bands. The grey dashed lines represent the flux limits of the 7 Ms CDF-S survey, i.e., FCDF−S = 6.4 × 10−18erg s−1cm−2 in the soft band, and
FCDF−S = 2.7 × 10−17erg s−1cm−2 in the hard band. About
40% (38%) of the HyLIRGs population shows a detectable FBH X
in the observed-frame soft (hard) band (hatched regions).
model (Magdziarz & Zdziarski 1995) in the XSPEC code12.
The metallicity and Eddington ratio λEdd = Lbol/LEdd are computed using the values predicted by GQd for each accret-ing BH.
12 The reflection component is computed assuming an isotropic
source located above the disc, fixing the reflection solid angle to 2π, the inclination angle to 60◦ and the reflection strength
parameter to R = 1, consistent with typical values of local AGNs (Zappacosta et al. 2018).
Figure 6. The blue histogram shows the distribution of NH
predicted for the overall population of galaxies in our model, while the grey and magenta histograms show the distribution of NH predicted for the ULIRG and HyLIRG progenitors of
the z ∼ 6 quasar host galaxies, respectively. The NH
distribu-tion for ULIRGs is consistent with the distribudistribu-tion of NH
pre-dicted for SMGs by the cosmological hydrodynamical simulation byNarayanan et al.(2015) (red hatched region).
We find that the BH activity in HyLIRGs generates an in-trinsic emission corresponding to X-ray luminosities (inte-grated over the rest-frame [0.5-10] keV band) of LBH
X =
[106.5 − 1012.5
] L (see Fig. 5.a), and to intrinsic fluxes FBH
X = [10 −20
− 10−14] erg s−1 cm−2 in the observed-frame soft/hard bands (see flux distributions in Fig.5.b and Fig.
5.c).
The intrinsic X-rays flux is attenuated by the interaction of the radiation produced during the BH accretion with the gas and dust in the immediate surroundings of the BH (mainly photoelectric absorption and Compton scattering of photons against free electrons). For this reason the effect of obscura-tion has to be taken into account, and the obscured emerg-ing flux can be written as Fabs
ν = Fνunabs e
−τν; at energies E & 0.1 keV, assuming a fully-ionized H-He mixture, the optical depth τν is (Yaqoob 1997):
τν= (1.2 σT+ σph) NH, (5) where NHis the hydrogen column density, while σTand σph are the Thomson and photoelectric cross section, respec-tively (seePezzulli et al. 2017for a description of σph and its dependence on energy and metallicity). We computed NHassuming the gas to be distributed following a singular isothermal sphere (SIS) profile (seeValiante et al. 2016), ρ(r) = ρ0
1 + ( r rcore)
2, (6)
where the normalization factor ρ0is a function of Mgas, the virial radius (Rvir) and the core radius (rcore).
By assuming rcore = 50 pc, we find that our simulated ULIRGs have NH= [1022− 1024] cm−2, in the same range predicted by cosmological hydrodynamical simulations for high-z SMGs (e.g.,Narayanan et al. 2015; see Fig.6). Adopt-ing the same rcore, we find that HyLIRG progenitors have NH∼ [1023− 1024] cm−2. As shown in Fig.5.b and Fig.5.c, the resulting net attenuation of fluxes by obscuration in the observed-frame soft X-ray band is about a factor of 2, while fluxes in the observed-frame hard X-ray band are almost unaffected by attenuation. This can be explained in terms of the photoelectric cross section decreasing for increasing
energy (see a thorough discussion inPezzulli et al. 2017). Finally, we compared the predicted, obscuration-attenuated FBH
X with the flux limits obtained by Chandra observa-tions of the recent 7 Ms Chandra Deep Field South Sur-vey (7 Ms CDF-S; Luo et al. 2017), i.e., FCDF−S = 6.4 × 10−18erg s−1 cm−2 in the soft band and FCDF−S = 2.7 × 10−17erg s−1 cm−2 in the hard band. We find that in both bands a significant fraction of our simulated HyLIRGs pop-ulation has a potentially detectable FXBH (40% and 38%, respectively; see hatched regions of the red histograms in Fig.5.b and Fig.5.c).
To test if FXcan be considered as a tracer of nuclear activity in high-z HyLIRGs, we estimated the contribution of SFR-related (e.g., X-ray binary populations; XRBs) processes to the X-ray luminosity, LSFRX , by using the redshift-dependent scaling relation between the rest-frame hard LX and SFR, found byLehmer et al.(2016):
log(LSFRX )[2−10] keV= 39.82+0.63 log(SFR)+1.31 log(1 + z). In Fig. 7, we show the relative contribution between the X-ray luminosity produced by the accreting SMBHs (once the attenuation by obscuration has been considered) and by XRBs, in the rest-frame hard band, as a function of the BH mass (and colour coded with the SFR), for the HyLIRGs progenitors with a detectable observed-frame soft13 FBH X (i.e., namely the objects located in the hatched region of the histograms in Fig.5.b). We find that HyLIRGs hosting black holes with MBH& 107 M have X-ray emission dom-inated by their SMBH activity. In particular, we find that, for about 40 % of the these galaxies, the X-ray luminosity due to the nuclear activity is at least one order of magnitude larger than that associated to star formation. Since we are conservatively assuming that the X-ray emission produced by stars is completely unobscured, if LSFR
X were corrected for attenuation14, the relative contribution of SMBH activ-ity to the total X-ray emission would be even higher. In conclusions, we find that (i) a significant frac-tion of HyLIRG progenitors has detectable (obscurafrac-tion- (obscuration-attenuated) X-ray fluxes produced by accreting BHs in both the observed-frame soft (40 %) and hard (38 %) bands; (ii) for those objects, the overall LX= LSFRX + LBHX is dominated by LBHX , making the observed X-ray emission an optimal tracer of SMBH activity. Combining these findings with the predicted spatial density of HyLIRGs (see Sec.3.1), we esti-mate the spatial density of HyLIRGs with detectable X-ray emission tracing SMBH activity to be ∼ 0.07 Gpc−3.
4 CONCLUSIONS
In the last decade, observations of DSFGs have extended to-ward z > 5. High-z DSFGs are often found to be HyLIRGs 13 The observed-frame [0.5-2] keV band traces hard X-ray
emis-sion at the redshifts of interest.
14
GQd model cannot provide information on the spatial distribu-tion of stars, therefore a self-consistent treatment of obscuradistribu-tion of the stellar X-ray emission is very difficult (seeValiante et al. 2016), without coupling detailed hydrodynamical models with ra-diative transfer simulations accounting for X-ray physics (Smidt et al. 2017;Graziani et al. 2018). On the other hand a statistical approach is currently impossible due to computational limits.
Figure 7. The relative contribution of rest-frame hard X-ray emission produced by nuclear activity and stars is shown as a function of MBH and colour coded depending on the SFR,
for galaxies with a detectable observed-frame soft X-ray flux (namely the objects in the red hatched region of Fig.5.b). At MBH & 107 M , the X-ray emission produced by accreting
SMBHs dominates over the X-ray emission produced by stars. For about 40% of the population, log(LBH
X /LSFRX ) > 1.
(LIR > 1013L ) with SFRs exceeding 1000M yr−1, likely tracing the result of strong dynamical interactions and in-tense gas accretion events occurring in the densest regions of the early Universe (Walter et al. 2012; Riechers et al. 2017; Pavesi et al. 2018). Recently, Marrone et al. (2018) found a IR-luminous system, SPT0311-58, at z ∼ 7, com-posed by a pair of extremely massive and IR-luminous inter-acting galaxies. Using different proxies to estimate the halo mass hosting the system, they show that SPS0311-58 marks an exceptional peak in the cosmic density field at this early cosmic time and it lies close to the exclusion curve predicted by the current structure formation paradigm.
In this work we use GQd model to simulate several indepen-dent merger histories of a typical luminous quasar at z ∼ 6, following the black hole growth and the baryonic evolution (including star formation, metals and dust life-cycles, and mechanical feedback) of its ancestors up to z ∼ 10. We find that:
- a fraction of progenitor galaxies (about 0.4 objects per single luminous quasar, between z ∼ 6 − 8) has HyLIRGs-like IR luminosity of LIR > 1013 L (see Fig. 2) and similar characteristics to the system observed byMarrone et al. (2018) (e.g., generally highly massive, extremely star-bursting and chemically evolved; see Fig.3);
- the IR-luminous progenitors of z ∼ 6 quasar host galaxies, reside in the most massive DM halos, with masses of MDM ∼ 1012.5 − 1013 M (see Fig. 1). Their far-IR continuum fluxes can be easily observed with ALMA with a modest amount of time, ranging from few seconds in Bands [8,7,6] (centered at observed-frame [0.7, 0.95, 1.25] mm, respectively) to about 2 minutes in Band 4 (2 mm) and 20 minutes in Band 3 (3 mm).
- the HyLIRGs progenitors of high-z quasar host galax-ies, have nuclear black holes with masses 105M
The IR-luminous progenitors of high-z quasars
9
108M
. The X-ray fluxes associated to the nuclear BH activ-ity could be detectable in the observed-frame soft and hard X-ray bands (they lie above the sensitivity limit of the 7 Ms Chandra Deep Field South Survey) for ∼ 40% of these galax-ies, even accounting for obscuration, and dominate over the X-ray emission associated to star formation.
Altogether our results suggest that z ∼ 6 luminous quasars are indeed the signposts of the observed rare high-z over-densities, and that massive-IR luminous galaxies at higher z are their natural ancestors. These findings corroborate mod-els of SMBH-galaxy co-evolution predicting an evolutionary link between IR-luminous galaxies and quasars.
ACKNOWLEDGEMENTS
The authors would like to thank D. Marrone for helpful dis-cussions. The research leading to these results has received funding from the European Research Council under the Eu-ropean Union’s Seventh Framework Programme (FP/2007-2013)/ERC Grant Agreement n. 306476. We have benefited from the public available programming language Python, in-cluding the numpy, matplotlib and scipy packages. This research made extensive use of ASTROPY, a community-developed core Python package for Astronomy (Astropy Col-laboration et al. 2013), and glueviz, a Python library for multidimensional data exploration (Beaumont et al. 2015).
REFERENCES
Alexander D. M., Hickox R. C., 2012,New Astronomy Reviews, 56, 93
Astropy Collaboration et al., 2013,A&A,558
Beaumont C., Goodman A., Greenfield P., 2015, in Taylor A. R., Rosolowsky E., eds, Astronomical Society of the Pacific Con-ference Series Vol. 495, Astronomical Data Analysis Software an Systems XXIV (ADASS XXIV). p. 101
Beelen A., Cox P., Benford D. J., Dowell C. D., Kov´acs A., Bertoldi F., Omont A., Carilli C. L., 2006,ApJ,642, 694 Bertoldi F., Carilli C. L., Cox P., Fan X., Strauss M. A., Beelen
A., Omont A., Zylka R., 2003,A&A,406, L55 Bianchi S., Schneider R., 2007,MNRAS,378, 973
Bischetti M., Maiolino R., Fiore S. C. F., Piconcelli E., Fluetsch A., 2018, preprint,p. arXiv:1806.00786(arXiv:1806.00786) Blain A. W., Smail I., Ivison R. J., Kneib J. P., Frayer D. T.,
2002,Phys. Rep.,369, 111
Brightman M., et al., 2013,MNRAS,433, 2485 Bromm V., 2013,Reports on Progress in Physics,76 Capak P. L., et al., 2011,Nature,470, 233
Carlstrom J. E., et al., 2011, Publications of the Astronomical Society of the Pacific,123, 568
Casey C. M., Narayanan D., Cooray A., 2014, Phys. Rep.,541, 45
Cattaneo A., Bernardi M., 2003,MNRAS,344, 45
Ceccarelli C., Viti S., Balucani N., Taquet V., 2018, MNRAS, 476, 1371
Chapman S. C., Blain A. W., Smail I., Ivison R. J., 2005,ApJ, 622, 772
Cicone C., et al., 2015,A&A,574
Coppin K. E. K., et al., 2008,MNRAS,389, 45
De Rosa G., Decarli R., Walter F., Fan X., Jiang L., Kurk J., Pasquali A., Rix H. W., 2011,ApJ,739, 56
Di Matteo T., Springel V., Hernquist L., 2005,Nature,433, 604
Dunne L., Eales S., Edmunds M., Ivison R., Alexander P., Clements D. L., 2000,MNRAS,315, 115
Fan X., et al., 2003,AJ,125, 1649 Fan X., et al., 2004,AJ,128, 515
Fan L., Knudsen K. K., Fogasy J., Drouart G., 2018,ApJ,856 Galliano F., et al., 2011,A&A,536, A88
Galliano F., Galametz M., Jones A. P., 2017, preprint, p. arXiv:1711.07434(arXiv:1711.07434)
Ginolfi M., Graziani L., Schneider R., Marassi S., Valiante R., Dell’Agli F., Ventura P., Hunt L. K., 2018, MNRAS, 473, 4538
Graziani L., Ciardi B., Glatzle M., 2018,MNRAS,479, 4320 Harrison I., Hotchkiss S., 2013,Journal of Cosmology and
Astro-Particle Physics,2013, 022
Harrison C. M., et al., 2012,MNRAS,426, 1073
Hayward C. C., Kereˇs D., Jonsson P., Narayanan D., Cox T. J., Hernquist L., 2011,ApJ,743
Hayward C. C., Narayanan D., Kereˇs D., Jonsson P., Hopkins P. F., Cox T. J., Hernquist L., 2013,MNRAS,428, 2529 Hopkins P. F., Hernquist L., Cox T. J., Di Matteo T., Robertson
B., Springel V., 2006,The Astrophysical Journal Supplement Series,163, 1
Hopkins P. F., Richards G. T., Hernquist L., 2007,ApJ,654, 731 Hopkins P. F., Hernquist L., Cox T. J., Kereˇs D., 2008,The
As-trophysical Journal Supplement Series,175 Hughes D. H., et al., 1998,Nature,394, 241
Ivison R. J., Papadopoulos P. P., Smail I., Greve T. R., Thomson A. P., Xilouris E. M., Chapman S. C., 2011, MNRAS,412, 1913
Jiang L., et al., 2016,ApJ,833, 222
Jones A. P., K¨ohler M., Ysard N., Bocchio M., Verstraete L., 2017,A&A,602, A46
Kov´acs A., Chapman S. C., Dowell C. D., Blain A. W., Ivison R. J., Smail I., Phillips T. G., 2006,ApJ,650, 592
Lacey C., Cole S., 1993,MNRAS,262, 627 Lehmer B. D., et al., 2016,ApJ,825, 7 Luo B., et al., 2017,ApJS,228, 2
Madau P., Dickinson M., 2014,Annual Review of Astronomy and Astrophysics,52, 415
Magdziarz P., Zdziarski A. A., 1995,MNRAS,273, 837 Maiolino R., et al., 2005,A&A,440, L51
Maiolino R., et al., 2012,MNRAS,425, L66 Marrone D. P., et al., 2018,Nature,553, 51
Murray S. G., Power C., Robotham A. S. G., 2013,Astronomy and Computing,3, 23
Narayanan D., et al., 2015,Nature,525, 496 Oteo I., et al., 2016,ApJ,827
Pavesi R., et al., 2018, preprint, (arXiv:1803.08048)
Pezzulli E., Valiante R., Schneider R., 2016,MNRAS,458, 3047 Pezzulli E., Valiante R., Orofino M. C., Schneider R., Gallerani
S., Sbarrato T., 2017,MNRAS,466, 2131
Richards G. T., et al., 2006,The Astrophysical Journal Supple-ment Series,166, 470
Riechers D. A., et al., 2011,ApJ,733 Riechers D. A., et al., 2013,Nature,496, 329 Riechers D. A., et al., 2017,ApJ,850
Robson I., Priddey R. S., Isaak K. G., McMahon R. G., 2004, MNRAS,351, L29
Romano-Diaz E., Shlosman I., Trenti M., Hoffman Y., 2011,ApJ, 736, 66
Salvadori S., Schneider R., Ferrara A., 2007,MNRAS,381, 647 Salvadori S., Ferrara A., Schneider R., 2008,MNRAS,386, 348 Sanders D. B., Soifer B. T., Elias J. H., Madore B. F., Matthews
K., Neugebauer G., Scoville N. Z., 1988a,ApJ,325, 74 Sanders D. B., Soifer B. T., Elias J. H., Neugebauer G., Matthews
K., 1988b,ApJ,328, L35
Sazonov S. Y., Ostriker J. P., Sunyaev R. A., 2004,MNRAS,347, 144
Silk J., Rees M. J., 1998, A&A,331, L1
Smail I., Ivison R. J., Blain A. W., 1997,ApJ,490, L5
Smidt J., Whalen D. J., Johnson J. L., Li H., 2017, preprint,p. arXiv:1703.00449(arXiv:1703.00449)
Springel V., et al., 2005,Nature,435, 629 Strandet M. L., et al., 2016,ApJ,822 Strandet M. L., et al., 2017,ApJ,842 Tanaka T., Haiman Z., 2009,ApJ,696, 1798
Valiante R., Schneider R., Salvadori S., Bianchi S., 2011,MNRAS, 416, 1916
Valiante R., Schneider R., Maiolino R., Salvadori S., Bianchi S., 2012,MNRAS,427, L60
Valiante R., Schneider R., Salvadori S., Gallerani S., 2014, MN-RAS,444, 2442
Valiante R., Schneider R., Volonteri M., Omukai K., 2016, MN-RAS,457, 3356
Volonteri M., Haardt F., Madau P., 2003,ApJ,582, 559 Walter F., Carilli C., Bertoldi F., Menten K., Cox P., Lo K. Y.,
Fan X., Strauss M. A., 2004,ApJ,615, L17 Walter F., et al., 2012,Nature,486, 233 Wang R., et al., 2008,ApJ,687, 848
Weingartner J. C., Draine B. T., 2001,ApJ,548, 296 Weiß A., et al., 2013,ApJ,767
Willott C. J., McLure R. J., Jarvis M. J., 2003,ApJ,587, L15 Willott C. J., et al., 2010,AJ,139, 906
Yaqoob T., 1997,ApJ,479, 184
Yue B., Ferrara A., Salvaterra R., Xu Y., Chen X., 2013,MNRAS, 433, 1556
Zappacosta L., et al., 2018,ApJ,854, 33
Zhukovska S., Dobbs C., Jenkins E. B., Klessen R. S., 2016,ApJ, 831, 147
da Cunha E., et al., 2013,ApJ,766, 13
de Bennassuti M., Schneider R., Valiante R., Salvadori S., 2014, MNRAS,445, 3039
This paper has been typeset from a TEX/LATEX file prepared by
